Elamur Research⌕
☰
home / papers / 2026 / EA-Agent: A Structured Multi-Step Reasoning Agent for Entity Alignment

EA-Agent: A Structured Multi-Step Reasoning Agent for Entity Alignment

Paper · 2026 · Yixuan Nan, Xixun Lin, Yanmin Shang et al.

Summary

Entity alignment (EA) means finding entities in two knowledge graphs that denote the same real-world object. Recent LLM-based EA systems give the LLM the final choice among candidates. This has two drawbacks: it is hard to see why a match was chosen, and long lists of triples make prompts large and costly. EA-Agent treats alignment as a planned, multi-step decision process. For each source entity, it first retrieves top-k candidates with TEA. It then plans a sequence of tool calls from a tool pool:

  • an Attribute Triple Selector and a Relation Triple Selector, which remove redundant triples before the LLM is called;
  • an Entity Alignment Tool;
  • an optional Reflector.

A reward scores both alignment correctness and path quality. The agent then rewrites weak paths and is fine-tuned offline with LoRA. On the FR–EN, JA–EN and ZH–EN benchmarks, the method beats earlier EA approaches and reports state-of-the-art results. It also uses far fewer tokens than comparable triple-based LLM methods. Because every decision follows an explicit tool path, errors are easier to trace and correct.

Key points

  • Evaluated on three cross-lingual benchmarks (FR–EN, JA–EN, ZH–EN), with gains of up to 3.17% in Hits@1 and consistent MRR improvements over prior SOTA.
  • Average prompt cost is 672 tokens per entity, much lower than ChatEA. Inference takes about 3.88 s per entity on FR–EN, versus 4.35 s for ChatEA and 0.89 s for Seg-Align, which ignores triples and loses accuracy.
  • Qwen3-32B is the strongest backbone. Llama-3-8B-Instruct slightly outperforms Qwen3-8B.
  • Accuracy improves over optimization rounds and levels off after the second round, so three rounds are used in all experiments.
  • Limitations: the tool set is designed by hand, and optimization relies on offline, reward-guided trajectory rewriting rather than end-to-end RL. Code is available at github.com/YXNan0110/EA-Agent.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🏛️Chinese Academy of SciencesBeijing · China194 works · 23,906 citations🎓University of Chinese Academy of SciencesBeijing · China147 works · 8,480 citations🔬Institute of Information EngineeringBeijing · China30 works · 1,411 citations🏢JingdongBeijing · China3 works · 870 citations🎓Donghua UniversityShanghai · China2 works

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 1,426 works

Cite

Cite as Yixuan Nan, Xixun Lin, Yanmin Shang, Ge Zhang, Zheng Fang, Fang Fang and Yanan Cao. 2026. EA-Agent: A Structured Multi-Step Reasoning Agent for Entity Alignment. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.1420

@inproceedings{nan2026ea,
  title = {{EA-Agent: A Structured Multi-Step Reasoning Agent for Entity Alignment}},
  author = {Yixuan Nan, Xixun Lin, Yanmin Shang, Ge Zhang, Zheng Fang, Fang Fang and Yanan Cao},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1420},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1420}
}

References

30 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating embeddings for modeling multi-relational data. Advances in neural information processing systems, 26.
  2. Muhao Chen, Yingtao Tian, Mohan Yang, and Carlo Zaniolo. 2016. Multilingual knowledge graph embeddings for cross-lingual knowledge alignment. arXiv preprint arXiv:1611.03954. find
  3. Xuan Chen, Tong Lu, and Zhichun Wang. 2024. Llm-align: Utilizing large language models for entity alignment in knowledge graphs. arXiv preprint arXiv:2412.04690. — [arXiv:2412.04690](https://arxiv.org/abs/2412.04690) find
  4. Yachao Cui, Hongli Yu, Xiaoxu Guo, Han Cao, and Lei Wang. 2024. Rakcr: Reviews sentiment-aware based knowledge graph convolutional networks for personalized recommendation. Expert Systems With Applications, 248:123403. find
  5. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, and 1 others. 2024. The llama 3 herd of models. arXiv e-prints, pages arXiv–2407.
  6. Lingbing Guo, Zequn Sun, and Wei Hu. 2019. Learning to exploit long-term relational dependencies in knowledge graphs. In International conference on machine learning, pages 2505–2514. PMLR. find
  7. Jinhao Jiang, Kun Zhou, Wayne Xin Zhao, Yang Song, Chen Zhu, Hengshu Zhu, and Ji-Rong Wen. 2025. Kg-agent: An efficient autonomous agent framework for complex reasoning over knowledge graph. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pa…
  8. Xuhui Jiang, Yinghan Shen, Zhichao Shi, Chengjin Xu, Wei Li, Zixuan Li, Jian Guo, Huawei Shen, and Yuanzhuo Wang. 2024a. Unlocking the power of large language models for entity alignment. arXiv preprint arXiv:2402.15048. find
  9. Yangqin Jiang, Yuhao Yang, Lianghao Xia, and Chao Huang. 2024b. Diffkg: Knowledge graph diffusion model for recommendation. In Proceedings of the 17th ACM international conference on web search and data mining, pages 313–321. find
  10. Chengjiang Li, Yixin Cao, Lei Hou, Jiaxin Shi, Juanzi Li, and Tat-Seng Chua. 2019. Semi-supervised entity alignment via joint knowledge embedding model and cross-graph model. Association for Computational Linguistics. find
  11. Zequn Sun, Wei Hu, Qingheng Zhang, and Yuzhong Qu. 2018. Bootstrapping entity alignment with knowledge graph embedding. In Ijcai, volume 18. find
  12. Xiaobin Tang, Jing Zhang, Bo Chen, Yang Yang, Hong Chen, and Cuiping Li. 2020. Bert-int: A bert-based interaction model for knowledge graph alignment. interactions, 100:e1. find
  13. Jiapu Wang, Kai Sun, Linhao Luo, Wei Wei, Yongli Hu, Alan W Liew, Shirui Pan, and Baocai Yin. 2024a. Large language models-guided dynamic adaptation for temporal knowledge graph reasoning. Advances in Neural Information Processing Systems, 37:8384– 8410. find
  14. Xixun Lin, Yanan Ning, Jing Zhang, and 1 others. 2025. Llm-based agents suffer from hallucinations: A survey of taxonomy, methods, and directions. arXiv preprint arXiv:2509.18970. — [arXiv:2509.18970](https://arxiv.org/abs/2509.18970) find
  15. Shijie Wang, Wenqi Fan, Yue Feng, Lin Shanru, Xinyu Ma, Shuaiqiang Wang, and Dawei Yin. 2025. Knowledge graph retrieval-augmented generation for llm-based recommendation. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 27152–…
  16. Xixun Lin, Hong Yang, Jia Wu, Chuan Zhou, and Bin Wang. 2019. Guiding cross-lingual entity alignment via adversarial knowledge embedding. In 2019 IEEE International conference on data mining (ICDM), pages 429–438. IEEE. find
  17. Yu Wang, Nedim Lipka, Ryan A Rossi, Alexa Siu, Ruiyi Zhang, and Tyler Derr. 2024b. Knowledge graph prompting for multi-document question answering. In Proceedings of the AAAI conference on artificial intelligence, volume 38, pages 19206–19214. find
  18. Yankai Lin, Zhiyuan Liu, Huanbo Luan, Maosong Sun, Siwei Rao, and Song Liu. 2015. Modeling relation paths for representation learning of knowledge bases. arXiv preprint arXiv:1506.00379.
  19. Zhichun Wang, Qingsong Lv, Xiaohan Lan, and Yu Zhang. 2018. Cross-lingual knowledge graph alignment via graph convolutional networks. In Proceedings of the 2018 conference on empirical methods in natural language processing, pages 349–357.
  20. Yu Liu, Yanan Cao, Xixun Lin, Yanmin Shang, Shi Wang, and Shirui Pan. 2025. Enhancing large language model for knowledge graph completion via structure-aware alignment-tuning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 20981–20995. find
  21. Yu Liu, Xixun Lin, Yanmin Shang, Yangxi Li, Shi Wang, and Yanan Cao. 2026. Pathmind: A retrieve-prioritize-reason framework for knowledge graph reasoning with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pages 15386–15393. find
  22. Apoorv Saxena, Adrian Kochsiek, and Rainer Gemulla. 2022. Sequence-to-sequence knowledge graph completion and question answering. arXiv preprint arXiv:2203.10321. find
  23. Jiashuo Sun, Chengjin Xu, Lumingyuan Tang, Saizhuo Wang, Chen Lin, Yeyun Gong, Lionel M Ni, Heung-Yeung Shum, and Jian Guo. 2023. Think-on-graph: Deep and responsible reasoning of large language model on knowledge graph. arXiv preprint arXiv:2307.07697. find
  24. Zequn Sun, Wei Hu, and Chengkai Li. 2017. Cross-lingual entity alignment via joint attribute-preserving embedding. In International semantic web conference, pages 628–644. Springer. find
  25. Yuting Wu, Xiao Liu, Yansong Feng, Zheng Wang, Rui Yan, and Dongyan Zhao. 2019a. Relation-aware entity alignment for heterogeneous knowledge graphs. arXiv preprint arXiv:1908.08210. find
  26. Yuting Wu, Xiao Liu, Yansong Feng, Zheng Wang, and Dongyan Zhao. 2019b. Jointly learning entity and relation representations for entity alignment. arXiv preprint arXiv:1909.09317. find
  27. Zhentao Xu, Mark Jerome Cruz, Matthew Guevara, Tie Wang, Manasi Deshpande, Xiaofeng Wang, and Zheng Li. 2024. Retrieval-augmented generation with knowledge graphs for customer service question answering. In Proceedings of the 47th international ACM SIGIR conference on research and development in inf… find
  28. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, and 1 others. 2025. Qwen3 technical report. arXiv preprint arXiv:2505.09388.
  29. Linyan Yang, Jingwei Cheng, and Fu Zhang. 2024a. Advancing cross-lingual entity alignment with large language models: Tailored sample segmentation and zero-shot prompts. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 8122–8138. find
  30. Linyao Yang, Hongyang Chen, Xiao Wang, Jing Yang, Fei-Yue Wang, and Han Liu. 2024b. Two heads are better than one: Integrating knowledge from knowledge graphs and large language models for entity alignment. arXiv preprint arXiv:2401.16960. — [arXiv:2401.16960](https://arxiv.org/abs/2401.16960) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/nan-2026-ea-agent-structured-multi-step/